# Géotechnique

Online ISSN: 0016-8505
Publications
Chapter
The sliding block model for the mechanism of deformation, in a body composed of grains, is based on the concept that movements of grains with respect to each other occur along planes that coincide preferably with the stress characteristic planes. In the case of plane strain these planes interest the two-dimensional plane of consideration along two characteristics lines called S1 and S2. The object of this Note is not to consider the probable veracity of such a model, but to establish the flow rule and the constitutive equations which follow from the special character of the model. The properties are taken to be those that were proposed by de Josselin de Jong (1958,1959). In that model sliding can occur simultaneously in the S1 and S2 directions at different shear strain rates, but limited in sense, and in addition the sliding elements are free to rotate.

Chapter
A cavity channel network consisting of many cavities with different compressibility interconnected by channels with different conductivity can serve as a model for a consolidating soil in both the primary and the secondary periods of consolidation. The abundance of the constituting elements is introduced as a continuous frequency function using the spring dashpot assembly as a model because it produces similar effects. It is shown how this frequency function can be determined from test results. Un réseau de cavités et de canaux comprenant de nombreuses cavités à compressibilités différentes reliées entre elles par des canaux à conductivités différentes peut servir de modéle pour un sol de consolidation aux deux périodes, primaire et secondaire de consolidation. L’abondance des éléments prenant part à la constitution est introduite comme fonction de fréquence continue en utilisant l’ ensemble à amortisseur comme modèle parce qu’il produit les mêmes résultats. On montre comment cette fonction de fréquence peut être établie à partir des résultats d’essais.

Article
Compacted expansive clays are often considered as a possible buffer material in high-level deep radioactive waste disposals. After the installation of waste canisters, the engineered clay barriers are subjected to thermo-hydro-mechanical actions in the form of water infiltration from the geological barrier, heat dissipation from the radioactive waste canisters, and stresses generated by clay swelling under almost confined conditions. The aim of the present work is to develop a constitutive model that is able to describe the behaviour of compacted expansive clays under these coupled thermo-hydro-mechanical actions. The proposed model is based on two existing models: one for the hydro-mechanical behaviour of compacted expansive clays and another for the thermo-mechanical behaviour of saturated clays. The elaborated model has been validated using the thermo-hydro-mechanical test results on the compacted MX80 bentonite. Comparison between the model prediction and the experimental data show that this model is able to reproduce the main features of volume changes: heating at constant suction and pressure induces either expansion or contraction; the mean yield stress changes with variations of suction or temperature.

Article
Extensive investigations have been and are being carried out on a stiff clay from an underground research laboratory located at Mol (Belgium) called Boom clay, in the context of research into deep nuclear waste disposal. Suction effects in deep Boom clay block samples were investigated through the characterisation of the water retention and of the swelling properties of the clay. The data obtained allowed an estimation of the sample initial suction that was reasonably compatible with the in-situ state of stress at a depth of 223 m. The relationship between suction and stress changes during loading and unloading sequences were also examined by running oedometer tests with suction measurements. A rather wide range of the ratio s/sigma 'v (being s the suction and sigma 'v the effective vertical stress) was obtained (0.61 - 1), different from that proposed by Bishop et al; (1974). Finally, the effect of suction release under an isotropic stress close to the estimated sample suction was investigated. A slight swelling (1.7%) was observed and further compression provided a satisfactory value of the overconsolidation ratio confirming the suggestions of taking some precautions before putting a swelling sample in contact with water as suggested by Graham et al. (1987). The various experimental data gathered in this study finally evidenced a relatively good state of conservation of the block sample used.

Article
The extraction of ores and minerals by underground mining often causes ground subsidence phenomena. In urban regions, these phenomena may induce small to severe damage to buildings. To evaluate this damage, several empirical and analytical methods have been developed in different countries. However, these methods are difficult to use and compare due to differences in the number of criteria used (from 1 to 12). Furthermore, the results provided by damage evaluation may be significantly different from one method to another. The present paper develops vulnerability functions based on a concept that has been applied in other areas, such as earthquake engineering, and that appears to be a more efficient way to assess building vulnerability in undermined cities. A methodology is described for calculating vulnerability functions in subsidence zones using empirical methods. The first part of the paper focuses on existing empirical methods for damage evaluation, and selected necessary improvements or modifications are justified. The second part focuses on the development of a building typology in subsidence zones and its application in the Lorraine region, where many villages are subject to subsidence problems due to iron-ore mining. The third section describes and discusses the adopted methodology for determining vulnerability and fragility functions or curves. Finally, vulnerability functions are tested and validated with a set of three subsidences that occurred in Lorraine between 1996 and 1999.

Chapter
The collapse height of a vertical cut off is computed by use of the variational calculus assuming the existence of a real slip line at collapse. A class of lilnes containing the real slip line is defined by total and local equilibrium conditions of the limit stress state. The extremal of the class is found to be an involute. Verification of the solution shows that the extremal gives either an unsafe estimate of the collapse height or corresponds to no extreme at all. These disappointing results are a consequence of the inadequate formulation of slope stability problems, when slip lines are computed by the calculus of variations. La hauteur, correspondant à la rupture d’un talus vertical, est déterminée à l’aide du calcul de variations en présupposant l’existence d’une ligne de glissement unique en cas de rupture. La classe de lignes, contenart cette ligne de glissement, est definie parl’équilibre total et local sous condition d’état de contraintes limites. L ’extrémale de la class possède la forme d’une involute. En vérifiant la solution, il est démontré que l’extrémale produit une hauteur de talus plus élevée que la hauteur de rupture, où une hauteur qui ne correspond pas du tout à un extremum. Ces résultats décevants sont engendrés par la formulation inepte des problémes de stabilité, quand des lignes de glissement sont déterminées à l’aide du calcul de variations.

Chapter
In this Paper it is shown how to use the double sliding, free rotating model for materials with internal friction to predict the stress history in undrained simple shear tests. In its original rigid plastic form this model could not be used, because there was no unique failure mode. By adding some elasticity to the prefailure stage (thus producing an elasto-plastic version of the model) this unique selection becomes possible. The extended model leads to explicit expressions for the stress history in a simple shear test. It is also shown how the failure mode taken by the model depends on the stress state at the start of the test. An active initial stress state leads to a ’toppling bookrow’ mode of failure, while a passive initial stress state produces horizontal sliding planes. With the exception of elasticity, the other properties of the double sliding model, including dilatancy, are taken in their original form. The essential features of the stress history obtained from the analysis resemble those actually observed in tests.

Article
A lunar soil simulant was used in research on predicting the performance of the Lunar Roving Vehicle (LRV) on the moon. The simulant was prepared from ground basaltic rock whose grain size distribution was matched to the lunar soil samples collected by Apollo 11 and 12. The strength characteristics of the simulant, i.e., internal friction angle, cohesion, and cone penetration resistance, were tested in triaxial tests, trenching tests, and cone penetration resistance tests. Subsequent soil tests and LRV performance on the moon proved that the lunar soil strength characteristics could be successfully simulated.-

Article
This paper summarizes our finding on the overturning potential of near-source ground motions. It is found that the toppling of smaller blocks is more sensitive to the peak ground acceleration; whereas the toppling of larger blocks depends on the incremental ground velocity. Introduction Early studies on the dynamic response of a rigid block supported on a base undergoing horizontal motion were presented by Housner (1963). In that study the base acceleration was represented by a rectangular or a half-sine pulse and expressions were derived for the minimum acceleration required to overturn the block. Using an energy approach he presented an approximate analysis on the dynamics of a rigid block subjected to a white noise excitation and he uncovered a scale effect that explained why the larger of two geometrically similar blocks can survive the excitation whereas the smaller block may topple. Yim et al. (1980) adopted a probabilistic approach and conducted a numerical study using artific...

Article
A new analytical method is introduced for calculating displacements due to tunnelling. This is conceived within the framework of the bound theorems of plasticity, but allowing for soil strain-hardening. The ground displacements due to tunnelling are idealised by a simple displacement mechanism of distributed shearing in the plane of the tunnel cross-section. The tunnel support pressure corresponding to a certain volume loss is calculated from energy balances of the work dissipated in distributed shear, the potential energy loss of soil flowing into the tunnel, and the work done by this soil against the tunnel support pressure. The calculations are carried out in steps of small volume loss accompanying small reduction in support pressure, after each of which the tunnel geometry is updated. In this way, each reduced tunnel support pressure is related to a complete ground displacement field. A simplified closed-form solution is also provided for the prediction of maximum surface ground settlement for the particular case of deep tunnelling. This closed-form solution is obtained by integrating the vertical equilibrium equation on the tunnel centreline from the tunnel crown up to the ground surface. These two analytical solutions have been validated against five previously published centrifuge tests.

Article
A kinematic plastic solution has been developed for ground movements around a shallow, unlined tunnel embedded within an undrained clay layer. In this solution, the pattern of deformation around the tunnel is idealised by a simple plastic deformation mechanism. Within the boundaries of the deformation mechanism, the soil is required to shear compatibly and continuously with no relative sliding at the boundaries. The soil is regarded here as a rigid plastic (Tresca) material. The plausibility of the proposed mechanism is demonstrated by comparison with limit analysis calculations and centrifuge test results.

Article
The unified three-dimensional critical state bounding-surface plasticity model for soil (gUTS) incorporates an anisotropic bounding surface by operating with a reduced deviatoric stress vector defined as the conventional deviatoric stress vector minus the dimensionless deviatoric vector {α} multiplied by the mean effective stress. Only when there is no Lode angle dependency will the deviatoric sections be geometrically similar and will the stress ratio corresponding to the critical state in the anisotropic case be the same as that in the isotropic case. If the effective stress ratio corresponding to the critical state depends on the reduced Lode angle, all deviatoric sections will remain geometrically similar for a given {α}, but the shapes of the sections will vary as {α} changes. These deviatoric sections generally become concave for large values of α. -Authors

Article
Synopsis A property of water in porous materials is that it freezes at temperatures below 0°C. There is no single freezing temperature for water in soils. As ice is formed the freezing point of the decreasing quantity of unfrozen water falls further below 0°C. Latent heat of fusion is thus involved in temperature changes over a range extending to several degrees below 0°C. The latent heat and specific heat together constitute an apparent specific heat. Apparent specific heats for various silt, clay and organic soils have been measured in a calorimeter. The apparent specific heats generally rise as temperatures approach 0°C, and in a clay soil may be ten times as great at −1°C as at −5°C. The apparent specific heats for a given temperature depend on whether the soil is freezing or thawing, and on various other factors. The precise determination of heat quantities involved in temperature changes in soil in situ is difficult, mainly because of the several factors influencing the freezing of the soil moisture. Une propriété de l'eau dans les corps poreux est qu'elle gèle à des températures inférieures à 0°C. Il n'y a pas de température unique de congélation de l'eau dans le sol. Au fur et à mesure que la glace se forme, le degré de congélation descend au dessous de 0°C. Une chaleur de fusion latente est done en cause lors des changements de température couvrant plusieurs degrés au dessous de 0°C. La chaleur latente et la chaleur spécifique constituent ensemble une chaleur apparente spécifique. Les chaleurs spécifiques apparentes pour des limons, des argiles et des sols organiques divers ont été mesurées dans un calorimètre. D'une manière générale, la chaleur spécifique apparente s'élève lorsque la température s'approche de 0°C, et dans le sol argileux elle peut être dix fois plus grande à −1°C qu'à −5°C. La chaleur spécifique apparente pour une température donnée dépend soit du gel, soit du dégel du sol, et d'autres facteurs divers. La détermination précise des quantités de chaleur en cause lors des channements de température dans le sol, sur place, est difficile, surtout à cause de plusieurs facteurs qui influencent la congélation de l'humidité se trouvant dans le sol.

Article
No The paper presents patterns and mechanisms of light non-aqueous phase liquid (LNAPL) migration in an unsaturated/saturated sand, based on a detailed experimental investigation using a fully instrumented two-dimensional model with dimensions of 120 x 120 x 10 cm. Suction head and degree of saturation were monitored simultaneously using tensiometers and time domain reflectometry (TDR) transducers respectively. LNAPL spills into the unsaturated zone were simulated to investigate the influence of new variables of practical importance, including the spill area, volume of spill and fluctuations of groundwater table, on the patterns of LNAPL migration. The patterns are explained in terms of the relationship between matric suction and degree of saturation. Fluctuations of water level are found to have a major influence on the distribution of LNAPLs in the unsaturated/saturated sand for large volumes of LNAPL spill. Measurements of degree of saturation of water at different levels are used to explain the LNAPL migration. It was found that water suction head was not affected by migrating LNAPL if the degree of saturation of water was above the residual saturation. Results of LNAPL suction head were found to be consistent with the migration patterns. Additionally, the average suction head difference between different levels indicated accurately the direction of LNAPL migration, which was in good agreement with the patterns observed using electronic imaging.

Article
The motion of the Aznalcóllar dam slide, after the initiation of the failure, is examined in the paper. The moving mass remained essentially rigid, and Newton’s second law provides the basic equation to determine the resulting dam displacement, velocity and acceleration. Resisting forces have been derived from a previous analysis of the conditions leading to failure. Pore water pressures induced by the forward motion of the dam are approximated as an undrained loading of the clay. Driving forcesand their evolution in time derive from the liquefaction of tailings and the condition of constant volume of liquefied tailings pushing the moving dam forward. The stable ‘cliff’ left by the motion inside the tailings lagoon supports this hypothesis. The model is able to reproduceclosely the distance travelled by the dam. It shows a lowsensitivity with respect to reasonable changes of the main parameters. The model computes that the slide motion lasted about 15 s, and that the maximum accelerationexperienced by the dam was 0.14g. Peer Reviewed

Article
Traditional railway foundations or substructures have become increasingly overloaded in recent years due to the introduction of faster and heavier trains. A lack of substructure re-engineering has resulted in maintenance cycles becoming more frequent and increasingly expensive. Two significant problems arising from increasing axle loads are differential track settlement and ballast degradation. One potential method of enhancing the substructure is by manipulating the level of ballast confinement. To investigate this possibility, a series of high frequency cyclic triaxial tests have been conducted to examine the effects of confining pressure and deviator stress magnitude on ballast deformation (permanent and resilient) and degradation. Experimental results indicate that for each deviator stress considered, an optimum range of confining pressures exists such that degradation is minimised. This range was found to vary from 15 – 65 kPa for a maximum deviator stress of 230 kPa, to 50 – 140 kPa when deviatoric stresses increase to 750 kPa. Ballast specimens tested at low confining pressures indicative of current in-situ conditions were characterised by excessive axial deformations, volumetric dilation and an unacceptable degree of degradation associated mainly with angular corner breakage. The results suggest that in-situ lateral pressures should be increased to counteract the axle loads of heavier trains, and practical methods of achieving increased confinement are suggested.

Article
Traditional railway foundations or substructures have become increasingly overloaded in recent years, owing to the introduction of faster and heavier trains. A lack of substructure re-engineering has resulted in maintenance cycles becoming more frequent and increasingly expensive. Two significant problems arising from increasing axle loads are differential track settlement and ballast degradation. One potential method of enhancing the substructure is to manipulate the level of ballast confinement. To investigate this possibility, a series of high-frequency cyclic triaxial tests has been conducted to examine the effects of confining pressure and deviator stress magnitude on ballast deformation (permanent and resilient) and degradation. Experimental results indicate that, for each deviator stress considered, an ?optimum? range of confining pressures exists such that degradation is minimised. This range was found to vary from 15– 65 kPa for a maximum deviator stress of 230 kPa to 50– 140 kPa when deviatoric stresses increase to 750 kPa. Ballast specimens tested at low confining pressures indicative of current in situ conditions were characterised by excessive axial deformations, volumetric dilation, and an unacceptable degree of degradation associated mainly with angular corner breakage. The results suggest that in situ lateral pressures should be increased to counteract the axle loads of heavier trains, and practical methods of achieving increased confinement are suggested.

Article
For foundations placed on horizontal ground surfaces, incorporating the effects of earthquake body forces, investigations have been performed by using the method of slices (Sarma & Iossifelis, 1990), limit equilibrium (Budhu & Al-Karni, 1993), the upper bound limit analysis (Richards et al., 1993; Dormieux & Pecker, 1995; Soubra, 1997, 1999) and the method of characteristics (Kumar & Mohan Rao, 2002a). For sloping ground surfaces, on the basis of limit analysis Zhu (2000) has reported an earthquake reduction factor for the bearing capacity factor,$N_{y}$. By using the method of characteristics, Kumar & Mohan Rao (2002b) have established the variation of the bearing capacity factors with changes in horizontal earthquake acceleration coefficient, h, for different ground inclinations. Two different failure mechanisms (single side and both sides) were used in the analysis. It was indicated that, in the presence of \alpha h, only the single-side failure mechanism remains statically admissible for the computation of Nc and Nq. By contrast, in finding$N_{y}$, for smaller values of \alpha h and _ both mechanisms were shown to remain statically admissible. As compared with the single-side mechanism, the bearing capacity factor $N_{y}$ was shown to become much smaller on the basis of the both-sides mechanism, especially for very small values of \alpha h and _. It was also indicated that the single-side mechanism remains kinematically admissible in all those cases where no slip has been allowed along the footing–soil interface (rough foundations). Although the approach was quite rigorous, however, it was assumed that the ratio of shear to normal stresses along the foundation–soil interface remains constant. Since the limit equilibrium technique does not require this assumption, it was intended to use this method to determine the seismic bearing capacity of rough foundations on sloping ground. It was simultaneously ensured that the solution remains kinematically admissible in all the cases. The results were thoroughly compared with those reported previously.

Article
An Underground Research Laboratory (URL) has been constructed in the deposit of Boom clay at a depth of 223 m by the SCK-CEN Belgian organisation near the city of Mol. This URL is devoted to research into nuclear waste disposal. This paper presents the results of an investigation carried out in the triaxial apparatus on specimens that were trimmed from blocks extracted during excavation sequences in the URL. In order to characterize the mechanical behaviour of the natural Boom clay and to examine the effect of initial and induced anisotropy on its constitutive behaviour, two series of triaxial tests were carried out. Special attention was devoted to the yield behaviour and the effects of stress history. Experimental results showed a clear relation between the shape of yield curves and stress history. The yield curve of the clay in its initial state was oriented along the K0 line, illustrating the anisotropy of fabric that was generated during the soil deposition. Subsequent isotropic compression (up to 9 MPa) made the yield curve more and more oriented along the p' axis. Based on these experimental results, an elasto-plastic model accounting for isotropic and anisotropic hardening in (p':q) space was developed. The initial yield curve of the soil was taken inclined with respect to the p' axis. The formulation proposed describes the change in shape, size and orientation of the yield curve, according to the stress history. Eight constitutive parameters were used to describe the anisotropic behaviour of the soil.

Article
The unsaturated shear strength parameters from two existing theories for two different compacted lateritic soils are presented. Fredlund, Morgenstern and Widger (1987) proposed a shear strength equation using two independent stress state variables, the net normal stress and the matrix suction. Toll (1990) put forward a framework for unsaturated soil behavior incorporating volume change as well. The parameters for Toll's theory were obtained using the peak, rather than the critical state, as the triaxial tests were carried out up to 10% axial strain and it was not possible to achieve critical state conditions within this strain level.

Article
Oedometer tests starting from a very small effective vertical stress of 0.5 kPa were performed on three reconstituted clays with different liquid limits. The soils were prepared at various initial water contents, ranging from 0.7 to 2.0 times their corresponding liquid limits. It is observed that the e-logσv´ compression curves show an inverse “S” shape due to suction pressure resisting deformation, similar to that of soft natural clays caused by consolidation yield stress. The suction pressure σs´ of the reconstituted clays can be correlated with the ratio of initial void ratio to void ratio at liquid limit e0/eL. The suction pressure curve (SPC) defined by a unique relationship between suction pressure σs´ and the normalised void ratio at suction pressure es/eL is also proposed to distinguish between the pre-suction and the post-suction states. In addition, Burland's concept of intrinsic compression line (ICL) is adopted for correlating the compression curves of various reconstituted clays at high initial water contents. It has been found that the void index is a powerful parameter for normalising the compression curves in the post-suction state. Nevertheless, it seems that Burland's ICL slightly underestimates the void index at the low stresses considered in this study. An extended intrinsic compression line (EICL) is then derived in order to better fit the data for stresses lower than 25 kPa.

Article
The present work aims at investigating the time-dependent behaviour of Boom clay, a potential host formation for the geological disposal of radioactive waste in Belgium, under both thermal and mechanical loading. High-pressure triaxial tests at controlled temperatures were carried out for this purpose. The results obtained confirmed the effect of the overconsolidation ratio on the thermal volume changes (thermal dilation under high OCRs and thermal contraction with OCR close to unity). Significant effects of temperature as well as of compression and heating rates were also observed on the volume change behaviour. After being loaded to a stress lower than the preconsolidation pressure (5MPa) at a low temperature of 25.8C and at a rate lower than 0.2kPa/min, the sample volume changes seemed to be quite small, suggesting a full dissipation of pore water pressure. By contrast, after being subjected to high loading and heating rates (including step loading or step heating), the volume changes appeared to be significant, particularly in the case of stresses much higher than the preconsolidation pressure. Because of its low permeability, full consolidation of Boom clay required a long period, and it was difficult to distinguish consolidation and creep from the total volume change with time.

Article
In this paper results from laboratory tests on London Clay and artificially cemented kaolin are presented and used to develop a preliminary framework for the time-dependent behaviour of soils, applicable to stiff clays and other soils. In the same way that the natural structure of clays has been shown to influence their monotonic behaviour, it is shown that it can also alter their response to changes in strain rate. The relative influence of the two main components of post-sedimentation structure - overconsolidation and diagenesis - on the time-dependent behaviour of London Clay was investigated in triaxial compression tests. The study was carried out in two steps, comparing first the behaviours of normally and overconsolidated reconstituted samples of London Clay subjected to stepwise changes in strain rate, and then the behaviours of overconsolidated reconstituted and undisturbed London Clay samples. The test results show that overconsolidation does not seem to affect the response of reconstituted London Clay to strain rate changes, which is consistent with published data on other stiff clays. However, intact and reconstituted overconsolidated samples show different behaviours, highlighting that it is the elements of structure resulting from diagenesis that influence the time-dependent behaviour of London Clay. Effects of cementing on strain rate sensitivity were investigated in triaxial compression of artificially cemented kaolin. The results obtained were different from the results for London Clay, suggesting that the difference in strain rate effects in the intact and reconstituted London Clay cannot be simply associated with cementing. A preliminary framework is proposed, where the time-dependent behaviour of soil depends on its particulate or continuum nature.

Article
The paper reports the results of an experimental study carried out on a bentonite compacted to a dry density ofup to 1·7 Mg=m3, a high value for this type of soil. The soil fabric has been studied using a variety of techniques, revealing a clear bimodal pore distribution that corresponds to two distinct structural levels: a microstructural one and a macrostructural one. The main testing programme has been performed using oedometers especially designed to apply a very large range of suctions. By applying the axis-translation technique (using nitrogen as the gas fluid), it has been possible to reach suctions up to 15 MPa. The higher suction range has been achieved byapplying a controlled atmosphere where the relative humidity has been fixed by a solution of sulphuric acid or salts. In this way suctions up to 550 MPa could be reached. The maximum vertical stress that could beapplied in the apparatus was 10 MPa. Two types of test have been carried out: (a) tests in which a combination of loading paths at constant suction and drying/wettingpaths at constant load were applied; (b) swelling tests under constant-volume conditions in order to determine the swelling pressure and the stress path followed during wetting. The results of the experimental programme are examined, taking into account the role of the soil fabricin controlling observed mechanical behaviour. In addition, the results of the laboratory tests are reproduced and interpreted using a generalised plasticity model that considers explicitly the interaction between macrostructure and microstructure. In this way, it is possible to achieve a more complete understanding of the mechanisms that underlie observed behaviour, and in particular the interplay between the two structural levels. Peer Reviewed

Article
This work shows that some soils with an intermediate behaviour between clays and sands do not have a unique compression line, as has also been pointed out by Martins et al. It has been shown here that a unique critical state line could not be determined for the remoulded soil. Although the results shown here are for a reconstituted residual soil, similar behaviour in compressibility has been found in silts and mixtures of sand and kaolin. However, it has been found that the critical state framework can be applied using a family of parallel critical state lines depending on the initial specific volume. One of the questions raised by this work is how the effects of structure can be quantified if the intrinsic properties of the remoulded material are not unique. The compression and critical state lines of the remoulded material are not only non-unique but also of a distinctly different gradient from the natural material.

Article
The paper presents an elasto-plastic model for unsaturated soils that takes explicitly into account the mechanisms with which suction affects mechanical behaviour as well as their dependence on degree of saturation. The proposed model is formulated in terms of two constitutive variables directly related to these suction mechanisms: the average skeleton stress, which includes the average fluid pressure acting on the soil pores, and an additional scalar constitutive variable, ξ, related to the magnitude of the bonding effect exerted by meniscus water at the inter-particle contacts. The formulation of the model in terms of variables closely related to specific behaviour mechanisms leads to a remarkable unification of experimental results of tests carried out with different suctions. The analysis of experimental isotropic compression data strongly suggests that the quotient between the void ratio, e, of an unsaturated soil and the void ratio es, corresponding to the saturated state at the same average soil sk...

Article
The behaviour of deep-ocean sediments from the North Atlantic abyssal plains is examined within the context of a recognised framework, the Sensitivity Framework, which was developed by Cotecchia & Chandler for structured terrestrial clays. The deep-ocean sediments have very high sensitivities, of the order of those of quick clays, but in volumetric compression their destructuration occurs at more moderate rates, similar to those usually found in terrestrial clays of medium sensitivities. The deep-sea sediments were retrieved using different types of sampler, which influenced greatly the strength and stiffness of the different samples. In particular, in the lower-quality samples the greater disturbance caused by sampling is responsible for changing their behaviour from brittle to ductile during shearing after compression to stresses post-gross yield, giving them an apparent insensitivity to shear strains. Results from numerical analyses using a constitutive model that was developed for structured terrestrial clays are presented. They show that such a model, which accounts for natural structure, can be used to simulate the behaviour of deep-ocean sediments provided an appropriate destructuration law is used in the model.

Article
This paper presents the results of an experimental study related to the thermal effects on kaolin clay cyclic mobility. The thermal effects were identified by comparing the experimental results of cyclic triaxial tests performed at high temperature (90 C) with results of the same type of test carried out at ambient temperature (22 C). For the testing, a new temperature-controlled triaxial apparatus, developed by the authors, was employed. Experimental evidence shows that shear cycles on the heated samples induced smaller axial strain and pore-water pressure per cycle in comparison with the unheated samples. In addition, shear-induced pore-water pressure at large strains in the heated sample was slightly lower than in the unheated sample. In other words, the heated samples behaved as if they were denser, which is a result of thermal hardening. These results may be applied in geotechnical and earthquake engineering applications as a soil improvement technique.

Article
This technical note presents a coupled thermo-hydro-mechanical model for unsaturated soil, which can be applied in the context of high temperature. A new pore gas or bulk air tranfer equation is introduced. Additionally the effect of temperature on the latent heat of vapourisation and specific heat capacities is incorporated. Adsorbed water is also included. The performance of the model is demonstrated via the simulation of a high temperature (150°C) therm-hydraulic-mechanical experiment carried out on a column of MX-80 bentonite by Commissariat à l'Energie Atomique, France. The model showed good correlation of experimental and numcerical results.

Article
Long-term strain records obtained in large-diameter oedometer tests on compacted gravels demonstrated that strains could be linearly related with the logarithm of time. Delayed compressibility coefficients were found proportional to the conventional stress-induced compressibility coefficients. A framework rooted on the phenomenon of crack propagation in rock particles induced by stress corrosion mechanisms is adopted to explain macroscopic observations. A model of crack propagation in loaded disc-shaped particles has been developed in order toexplain the nature of particle breakage and its relationshipwith time, macroscopic stress and total suction.Experimental observations such as the existence ofthreshold stresses that mark the onset of delayed deformations are explained by the model. It was also foundthat the main features of the delayed deformation ofrockfill could be physically explained within the developed framework. In particular, a simple closed-form relationship between the coefficient of delayed deformation, the compressibility coefficient and a parameter describing the rate of crack propagation could be found.It was found to be consistent with experimental observations. Peer Reviewed

Article
Thesis (doctoral)--University of Cambridge, 1974.

Article
"Crampton Price" de la Institution of Civil Engineers (London) Beliche Dam, a zoned earthdam with rockfill shouldersand a central clay core, experienced large collapse settlementsdue to reservoir impounding and direct action of rainfall. Long-term field records of vertical and horizontal displacements are available as well as a set of largescalelaboratory tests on rockfill specimens. It has been the subject of several numerical analyses that failed to capture the relevant effect of weather conditions on thebehaviour of the dam. Recent developments in the constitutive modelling of rockfill allow a substantial improvement of modelling capabilities, and this is illustrated inthe paper. Laboratory test results under dry and flooded conditions were interpreted, and material parameters were identified. The complete history of dam construction,impoundment and rainfall was then simulated by means of a coupled flow–deformation model. Deformations during construction and impoundment have been reproduced. Long-term deformations have consistently been related to rainfall records. In general, long-termdeformations are controlled by the varying wetting history of the dam shoulders and by an intrinsic deformation component. The wetting action comes to an end when the relative humidity of the rockfill reaches 100% for the first time. The paper also discusses scale effects and the role of rockfill permeability in the developmentof deformations. Peer Reviewed

Article
Introduction: Bender elements were first introduced to soil testing by Shirley & Hampton (1978), and their use has now became almost routine in a wide variety of apparatus, using designs similar to those described in detail by Dyvik & Madshus (1985). Benders transmit and receive S-waves through the soil, enabling the shear wave velocity to be determined and hence the shear modulus, an important geotechnical parameter. Shear plates are also used, which are more effective at high confining stresses, and with very coarse soils (Brignoli et al., 1996). However, they are bigger, require much larger driving voltages, and at low confining stresses are less efficient than benders, owing to the large mismatch between the characteristic impedance of the element and the soil (Shirley & Hampton, 1978). Other flat piezoelectric transducers have been used to transmit and receive P-waves through the soil, enabling the compression wave velocity to be determined and hence the constrained modulus. Various authors have described experimental set-ups that incorporate both S-wave and P-wave transducers mounted in the same apparatus (e.g. Schultheiss, 1981; Bates, 1989; Nakagawa et al., 1996; Brignoli et al., 1996). This Note explains how a very simple modification to a standard bender element design can result in a transducer capable of transmitting and receiving P-waves; such transducers are referred to here as extenders. A further slight change to the wiring configuration leads to a single hybrid element termed a bender/extender, capable of transmitting and receiving both Sand P-waves. The effectiveness of the combined bender/extender elements is demonstrated using test data acquired in a 100 mm dia. triaxial sample of dry Leighton Buzzard sand.

Article
Synopsis In order to instrument a deep clay deposit for measurements of consolidation beneath a 23-ft-high embankment, it was necessary to develop a settlement gauge which could be installed just below the natural ground surface before the embankment was constructed. The Paper presents the considerations involved in the design of such a gauge and describes the one finally adopted. Measurements of settlements between the natural ground surface and reference elevations within the clay deposit are made by determining electrical resistances which are proportional to vertical displacements. This is accomplished by the simple gearing of a chain and sprocket to the shaft of a spirally wound potentiometer. By a judicious choice of potentiometer and sprocket pitch diameter it is possible to determine displacements to an accuracy of 0·05 in or 1·3 mm. The field installation is described in detail and, finally, time-settlement curves for four gauges are shown. Afin de mesurer scientifiquement le tassement d'un profond dépôt d'argile situé sous un remblai mesurant 23 pieds de haut il a fallu concevoir une jauge spéciale qu'on pourrait installer sous la surface du terrain naturel avant de faire le remblai. On passe en revue les considérations dont on a term compte dans la mise au point de ce type d'instrument et on décrit la jauge qu'on a finalement adoptée. La mesure des tassements produits entre la surface du sol naturel et les points de référence choisis à différentes hauteurs dans le dépôt s'est effectuée en déterminant les résistances électriques lesquelles sont proportionnelles aux déplacements verticaux. On a réalisé cette mesure en engrenant simplement une galle à l'axe d'un potentiomêtre hélicoïde. En choisissant comme il faut le diamètre du pas on a pu déterminer les déplacements à 0·05 pouce ou 1·3 mm près. L'installation utilisée pour la mesure est décrite en détails et l'étude est complétée par des courbes de tassement établies en fonction du temps.

Article
A modification to Rowe’s stress-dilatancy equation is presented that extends its range of application to include unsaturated soil behaviour. The results of a programme of constant water content triaxial tests on unsaturated bentonite-enhanced sand (BES) are reported, together with those of a programme of saturated drained triaxial tests on the sand. It is shown that the variation in the rate of dilation at failure with the sand relative density is similar for the two materials. It is proposed that the packing and friction angle of the sand particles and the degree of saturation control the shear strength of unsaturated BES containing modest amounts of bentonite, and that the shear strength of the bentonite component can be ignored.

Article
INTRODUCTION Barriers with a low hydraulic conductivity are used as part of waste containment systems to prevent groundwater contamination by liquids from the waste. Commonly barriers are either a geomembrane (usually an HDPE sheet), a mineral layer or a combination of the two. Recently there has been increasing interest in the use of bentonite±sand mixtures as the mineral layer, in both land®ll liners and vertical cut-off walls, partly because they are less susceptible to frost damage and desiccation cracking than compacted clay (Dixon et al., 1985; Kraus et al., 1997). Currently there is uncertainty about the strength and bearing capacity of these materials. This note reports drained strength data for bentonite±sand mixtures and proposes that trends in these data are mainly the result of variations in the relative density of the sand.

Article
The research and development programme on geological disposal for high-level and long-lived waste (HLW) in Belgium was initiated in 1974. A deep tertiary clay formation, the Boom Clay, present under the Mol-Dessel nuclear site, was selected as a reference host formation for experimental purposes. The construction of the underground laboratory HADES (at a depth of 223 m, initiated in 1980 and extended in 2002) allowed the building of a valuable geotechnical database and led to the development of improved excavation techniques that significantly reduce the Excavation Damaged Zone (EDZ). Since the operational start of HADES about 25 years ago, many geotechnical measurements have been performed around excavations. Comparison between in-situ measurements and modelling results allowed a continuous improvement of our knowledge on the Boom Clay behaviour. Important issues to interpret correctly the measurements are the good control of the excavation parameters and the boundary conditions. An extensive characterisation of the hydromechanical response of Boom Clay around an excavation for short and long term conditions has been performed. One important finding was the occurrence of measurable hydraulic effects at a distance about 60 m (12.5 tunnel diameters) ahead of the tunnel excavation.

Article

Article
A study has been carried out on the effect of stress path geometry on drained soil brittleness as measured in the laboratory. It shows that an explicit relationship exists between brittleness and stress path geometry for soils of given peak and post-peak strength envelopes. For the particular case of triaxial testing, four typical effective stress paths of shearing have been considered. The results indicate that the conventional drained triaxial test yields the highest brittleness, either when the stress paths are reaching failure at the same confining stress or when they start at the same consolidation pressure prior to shearing. The practical implications of the study are discussed. Une étude a été réalisée sur l'effet de la géométrie du chemin des contraintes sur la fragilité de sols drainés, tel que mesuré en laboratoire. Cette étude montre qu'il existe un rapport explicite entre la fragilité et la géométrie du chemin des contraintes pour des sols à résistances de pic et residuelle données. En ce qui concerne le cas particulier d'essais triaxiaux, on a considéré quatre chemin de contrainte efficaces typiques de cisaillement. Les résultats montrent que c'est avec I'essai triaxial drainé classique qu'on obtient la plus haute fragilité, que les chemins de contrainte atteignent la rupture à la même tension de confinement ou qu'ils commencent à la même pression de consolidation avant cisaillement. Les implications pratiques sont analysites dans cette étude.

Article
Granular filters are used in earth structures, such as embankment dams, to protect fine soils from erosion due to seepage forces. Successful filtration requires that the filter voids are fine enough to capture some of the coarse fraction of the base soil. These retained particles are then able to capture progressively finer base soil particles, and eventually a filter interface forms that is able to prevent any further erosion. This process is called self-filtration. Lafleur et al. (1989) examined self-filtration in cohesionless, broadly graded base soils. It was found that the extent of mass loss before selffiltration occurs was greater in broadly graded materials: hence a finer filter was required to reduce this mass loss. Filters for cohesive base soils are commonly designed using the Sherard & Dunnigan (1985) design criteria. While these criteria have been developed from extensive laboratory data, they may not be applicable to all fine base soils, particularly broadly graded materials. In this paper, a series of filtration tests on various base soils are described. Data from the current study, and the published results of laboratory tests from several sources are compared to examine the filtration of broadly graded base soils. Based on this analysis, a new design procedure is proposed for filters to protect fine base soils, which determines the ability of the coarse fraction of the base soil to retain the fine fraction (i.e. a self-filtering base soil).

Article
There are many methods for the analysis and design of embedded cantilever retaining walls. They involve various different simplifications of the net pressure distribution to allow calculation of the critical retained height. In the UK, it is commonly assumed that net pressure consists of the sum of the active and passive limiting pressure values. In the USA, the net pressure is commonly simplified by a three-line rectilinear pressure distribution. Recently, centrifuge tests have led to a proposed semi-empirical rectilinear method in which an empirical constant defines the point of zero net pressure. Finite element analyses presented in this paper examine the net pressure distribution at limiting equilibrium. The study shows that the point of zero net pressure for a best-fit rectilinear approximation is dependent on the ratio between the passive and active earth pressure distributions at limiting conditions. A simple empirical equation is proposed which defines the point of zero pressure. The predictions for the critical retained height and bending moment distribution using this empirical equation are in excellent agreement with the finite element results and centrifuge data. They are in better agreement than the predictions of the commonly used analysis methods.

Article
The trend toward physical modelling of larger prototypes triggers a move towards expansion of the current capability of dynamic centrifuge tests. In this paper a method based on the concept of two-stage scaling is proposed that allows expanded use of the currently available facilities. In the first stage a prototype is scaled down into an intermediate virtual model based on the scaling relations in a 1g field with a scaling factor of μ (prototype/virtual model). In the second stage this model is scaled down into a physical model using the conventional scaling relations in the centrifugal field with a scaling factor of η (virtual model/physical model). In this manner, a large scaling factor λ = μη (prototype/physical model) can be broken down into the smaller scaling factors μ and η, and thus the centrifuge model tests can be performed with the smaller scaling factor η. In effect, the limits in the scaling factor for the currently available centrifuge facilities can be expanded by a factor of μ. Validation of the proposed method is provided by algebraic derivations based on the mechanics of model tests.

Article
A kinematic plastic solution has been developed for the penetration of a circular footing into an incompressible soil bed. In this solution, the pattern of deformation around the footing is idealised by a simple plastic deformation mechanism. Strain-hardening behaviour and nonlinear stress–strain characteristics are incorporated. This application is different from conventional applications of plasticity theory as it can approximately predict both stresses and displacements under working conditions. This approach therefore provides a unified solution for design problems in which both serviceability and safety requirements are based directly on the stress–strain behaviour of the soil. The design strength that should limit the deformations can be selected from the actual stress– strain data recorded from a carefully specified location, and not derived using empirical safety factors. The validity of this design approach is examined against nonlinear finite element analyses and field measurements of foundations on clay under short-term loading.

Article
Piezocone penetration tests were carried out at five sites in marine clay with overconsolidation ratios ranging from 1·3 to 4·8 A simple model for the behaviour of these clays indicates that the effective vertical yield stress mobilized along the cone axis during penetration is equal to the difference between the actual induced total vertical stress in the soil and the total porewater pressure. The actual induced total vertical stress can he derived from piezocone test data. Theoretically the ratio of the vertical yield stress inferred from piezocone results and the preconsolidation pressure should correlate with the overconsolidation ratio. This hypothesis is supported by the field and laboratory investigations presented in this study. Des essais de pénétration avec un piezocone ont été effecté sur cinq sites d'argile marine ayant des rapports de surconsolidation variant de 1,3 à 4,8. Un modèle simple de comportement de ces argiles indique que la contrainte effective verticale à l'état limite mobilisée durant la pénétration est égale à la différence entre la contrainte totale verticale et la pression interstitielle générées dans le sol. La contrainte totale verticale peut être déterminée à partir des réultats des essais de piezocone. En théorie, le rapport de la contrainte verticale à l'état limite, calculée à partir des résultats du piezocone, et de la pression de surconsolidation devrait être une fonction du rapport du surconsolidation. Cette hypothése est vérifiée par les essais in situ et de laboratoire présentés dans cet article.

Article
Joints existing within a rock mass are normally filled with fine materials such as clay and silt, which decrease its ultimate strength and make it easier to deform. The shear behaviour of infilled rock joints is controlled by several parameters, such as infill thickness, joint roughness and drainage conditions (Ladanyi & Archambault, 1977; Lama, 1978; de Toledo & de Freitas, 1993; Indraratna et al., 2005). If the infill material is saturated and drainage is impeded, pore water pressure within the infilled joint will build up and change during shear displacement. The importance of pore water pressure in the hydromechanical stability of jointed rock mass has been recognised for several decades (Lane, 1970). Several catastrophic failures of natural rock slopes have occurred as a result of pore water pressure build-up in jointed rock (e.g. Kangaroo Valley rockslide, NSW, Australia; Indraratna & Ranjith, 2001). Limited experimental studies on the effect of pore water pressure in rock specimens (porous or fractured) have been reported by Lane (1970), Goodman & Ohnishi (1973) and Poirier et al. (1994). Even fewer studies have been dedicated to jointed rock behaviour under undrained conditions including pore water pressure behaviour, but without any infill (e.g. Archambault et al., 1998, 1999). In this paper, the authors describe some limited but interesting test results on the behaviour of clay-infilled joints sheared under triaxial, undrained conditions.

Article
The effect of the temperature dependence of the internal friction angle is studied in a boundary value problem simulating the impact of a cylindrical heat source on the soil mass in which it is embedded. This follows a previous study which shows that such temperature dependence may substantially affect the interpretation of thermal failure in laboratory experiments. Even if the thermal increase of the internal friction is quite modest (less than 20% in terms of the critical state parameter, M), it affects quite significantly the effective stress path near the heat source. The effective stress path approaches the yield locus and the critical state at significantly higher principal stress difference values for the variable internal friction than for the M const case. The ‘mean effective stress distance from the critical state’ is substantially reduced during heating, which in the case of small perturbations of any parameter may lead to potentially unstable or statically inadmissible behaviour. The solutions obtained allow one to identify zones of influence around the heat source of several variables of interest. The two fields most affected by the thermal sensitivity of M are that of the axial stress, dropping significantly near the heat source, and that of the appearance of the thermoplastic strain. Both zones of influence are reduced in size by almost half when the friction angle is increased by 20% over 708. The presented results may be of relevance to the design of prototype in situ installations and their monitoring, and eventually of actual facilities for nuclear waste disposal.

Article
The true cohesion of an undisturbed, sensitive clay was measured by compression testing of laboratory specimens at very low effective stresses. On the basis of modern ideas, the measured cohesion may be visualized as a frictional resistance mobilized by the arrangement of, and the intrinsic stresses between, individual soil particles. This concept, supported by observed fracture patterns in the laboratory and in the field, is considered to lead to a better fundamental understanding of the strength of undisturbed clay. Extrapolation of strength theories developed for remoulded clays to the assessment of undisturbed, sensitive clays is refuted. La cohésion vraie d'une argile intacte et sensible fut mesurée au moyen d'essais de compression, á tension efficace trés basse, sur des échantillons de laboratoire. Selon les idées contemporaines, la cohésion mesurée peut étre comparé á une friction mobiliske par l'arrangement des particules individuelles et les tensions intrinséques entre elles. Ce concept soutenu par le genre de ruptures observées in situ et en laboratoire, conduit á une meilleure compréhension fondamentale de la résistance d'une argile intacte. L'extrapolation des théories de résistance pour les argiles remaniées dans le domaine des argiles intactes et sensibles est réfutée.

Article
An investigation into the pore water pressure regimes in the actively eroding London Clay coastal slopes at Warden Point, Isle of Sheppey, Kent included the installation of 56 standpipc piezometers at a range of depths down to 60 m below original ground level. As expected, given the active coastal environment, a significant number of the instruments were destroyed before final equilibration water levels had been obtained. Using initial equilibration data from the fully equilibrated instruments, it is demonstrated that Hvorslev's theory for the response of a piezometer located in incompressible soil is valid for standpipe piezometers installed in unweathered London Clay. Using this theory a simple method is proposed for predicting the final equilibration water level of partially equilibrated instruments. The initial equilibration data from 13 of the piezometers were of sufficient quality to enable calculations of in situ permeability. These demonstrate a relationship of decreasing permeability with depth for London Clay, conforming to and greatly extending the small number of existing results.

Top-cited authors
• University of Minnesota Twin Cities
• University of Cambridge
• Universitat Politècnica de Catalunya
• University of Oxford
• Imperial College London